Sequence Pattern Mining based on Markov Chain

Sequence pattern mining is one of the main challenges in data mining and especially in large biological sequence databases, which consist of a large number of DNA sequences. Many existing methods are time consuming and scan the database multiple times. In order to overcome such shortcomings, we propose a fast and efficient algorithm SPMM based on Markov chain for mining sequence patterns because the DNA sequences meet Markov property. We first present the relative concepts and definitions. And then SPMM algorithm is put forward in which transition probabilities matrix is computed for each DNA sequence. The sequence patterns can be identified according to the given threshold of minimum support degree. Some examples are given to illustrate SPMM in detail. The experimental results show that our SPMM algorithm can achieve not only faster speed, but also higher quality results as compared with other algorithms.